Discovery of Novel CXCR2 Inhibitors Using Ligand ... - ACS Publications

Page 1 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Chemical Information and Modeling

(Journal of Chemical Information and Modeling)

Discovery of Novel CXCR2 Inhibitors Using Ligand-Based Pharmacophore Models

Helen Ha†, Bikash Debnath‡, Srinivas Odde†, Tim Bensman§, ¶, Henry Ho§, ¶, Paul M. Beringer§, and Nouri Neamati‡,* †

Department of Pharmacology and Pharmaceutical Sciences, University of Southern California,

School of Pharmacy, 1985 Zonal Avenue, Los Angeles, California 90033, United States ‡

Department of Medicinal Chemistry, College of Pharmacy, and Translational Oncology

Program, University of Michigan, North Campus Research Complex, 2800 Plymouth Road, Ann Arbor, Michigan 48109, United States §

Department of Clinical Pharmacy and Pharmaceutical Economics & Policy, University of

Southern California, School of Pharmacy, 1985 Zonal Avenue, Los Angeles, California 90033, United States ¶

Authors contributed equally

*Corresponding Author: Email: [email protected]. Phone: 734-647-2732

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT: The chemokine receptor, CXCR2 is expressed on various immune cells and is essential for neutrophil recruitment and angiogenesis at sites of acute and chronic inflammation caused by tissue injury or infection. CXCR2 and its ligand, CXCL8, are implicated in a number of inflammation-mediated diseases such as COPD, cystic fibrosis, and cancer. Though the development of CXCR2-specific small-molecule inhibitors as anti-inflammatory agents has been pursued by pharmaceutical companies within the past decade, there are currently no clinically approved CXCR2 inhibitors. A pharmacophore model based on previously reported CXCR2 antagonists was developed to screen a database of commercially available compounds. Smallmolecule compounds identified from the pharmacophore screening were selected for in vitro screening in a cell-based CXCR2-mediated β-arrestin-2 recruitment assay and further characterized in several cell-based assays and LPS-induced lung inflammation studies in mice. CX compounds identified from pharmacophore modeling inhibited cell migration, lung and colon cancer cell proliferation and colony formation. Mechanistic studies of CX4152 show that this compound inhibits CXCR2 signaling through down-regulation of surface CXCR2. Additionally, CX4152 significantly inhibited CXCL8-mediated neutrophil migration and LPSinduced lung inflammation in mice. Using a CXCR2 inhibitor-based pharmacophore model, we have identified a novel set of sulfonamides from a diverse library of small-molecules. These compounds inhibit CXCR2/β-arrestin-2 association, cell migration and proliferation, as well as acute inflammation in mouse models. CX compounds identified from our pharmacophore models are potential leads for further optimization and development as anti-inflammatory and anticancer agents.

2


Page 2 of 70

Page 3 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


KEYWORDS: Pharmacophore modeling, CXCR2, CXCL8, inflammation, cancer, smallmolecule compounds

3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

INTRODUCTION The CXC chemokine receptor 2 (CXCR2) is a G-protein coupled receptor (GPCR) expressed on inflammatory cells such as neutrophils, monocytes, T-lymphocytes and basophils as well as endothelial cells and CNS neurons.1-5

CXCR2 is activated by several ELR+ chemokines,

including, interleukin-8 (IL-8 or CXCL8), epithelial cell-derived neutrophil-activating protein 78 (ENA-78), neutrophil-activating peptide 2 (NAP-2), and Groα.1, 6 The CXCR2 ligand, CXCL8, is essential for neutrophil chemotaxis and for angiogenesis during acute and chronic inflammatory conditions caused by tissue injury or infections.7-10 It also activates CXCR1, another chemokine receptor that is also involved in neutrophil chemotaxis.1 Since CXCR2/CXCL8 has a critical role in inflammation, CXCR2 inhibition may be able to alleviate symptoms of respiratory diseases characterized by neutrophil-mediated inflammation, such as chronic obstructive pulmonary disease (COPD) and cystic fibrosis (CF). Clinical studies show that airway secretions and lung tissue from patients with COPD have increased CXCR2 chemokine ligand levels compared to healthy patients, which negatively correlates with lung function.11,

12

Realizing the anti-inflammatory potential of CXCR2 inhibition, several

pharmaceutical companies began to pursue the development of small-molecule CXCR2 inhibitors using high-throughput screening platforms. The first class of CXCR2 inhibitors was discovered in the mid 1990s by GlaxoSmithKline. SB225002 was the first reported CXCR2selective small-molecule compound, and further developments led to SB265610 and SB656933 (Table 3).13, 14 CXCR2 is also expressed by a number of different cancer cells, including lung, colon, ovarian, and prostate cancer cells.15-18 CXCL8 derived from tumor-associated macrophages, neutrophils and tumor cells promotes cancer cell proliferation, survival, invasiveness, tumor angiogenesis 4


Page 4 of 70

Page 5 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and decreased sensitivity to anticancer agents.19-21

In addition to CXCL8, other CXCR2

chemokine ligands such as Groα also significantly enhance proliferation and invasiveness of cancer cells.22, 23 Lung cancer xenograft studies show that CXCR2 knockout mice have reduced tumor size, reduced spontaneous metastases, and decreased vascular density compared to wildtype mice. In addition, CXCR2+/+ mice treated with CXCR2 neutralizing antibodies exhibit phenotypes similar to CXCR2-/- mice.24 These preclinical studies indicate that CXCR2/CXCL8 is an essential component of the tumor microenvironment, promoting tumor progression, and its inhibition may be therapeutically beneficial for cancer patients. However, currently, there are no approved CXCR2 inhibitors, nor are any in clinical trials to assess their efficacy in cancer patients. Given the anti-inflammatory and anticancer therapeutic potential of CXCR2 inhibition, we sought to develop a robust pharmacophore model to identify novel classes of compounds that inhibit CXCR2 function as well as show anti-inflammatory and anticancer properties in in vitro and in vivo models.

MATERIALS AND METHODS Molecular Modeling. Molecular modeling was performed on a Dell Precision T7400n MiniTower, Quad Core Xeon Proc X5450 dual processor with 32 nodes computer. Catalyst software (version 4.11, Accelrys Inc., San Diego, CA) was used to generate HypoGen pharmacophore models.25 Molecular docking was performed using GOLD software v 5.2 (Cambridge Crystallographic Data Centre (CCDC), UK).26, 27 ROCS 3.2.0.4 (OpenEye Scientific) software was used to generate shape-based pharmacophore models.28, 29 Data Set and Modeling of HypoGen Pharmacophore. Biological activity data (binding assay performed with Chinese hamster ovary (CHO) cell lines expressing CXCR2), represented as IC50 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 70

(nM) were obtained from literature.30-35 Chemical structures of selected CXCR2 antagonists were separated into training and test set based on activity and structural diversity (Supplementary Tables 1 and 2, respectively). Three different chemical classes of CXCR2 inhibitor (N,N'diarylsquaramides, N,N'-diarylureas, and N,N'-diarylguanidines) were chosen for pharmacophore hypothesis generation (Figure 1). Both training and test sets have a structurally similar set of compounds and representatives from the 3 different classes to ensure structural diversity. The activities (IC50) against CXCR2 are classified as: highly active (1 µM, +). Both training and test set have a uniform distribution of compounds with these three activity classes. Our training set consisted of 20 compounds with significant structural diversity and wide range of bioactivity with IC50 ranging from 4 nM to 25 µM.

Structures of compounds of a specified configuration are drawn using

ChemDraw v12 (PerkinElmer, Waltham, MA) and energy minimization was performed using Open Force Field (OFF) methods with the steepest descent algorithm within Discovery Studio 2.0 (Accelrys Inc., San Diego, CA).36 A gradient convergence value of 0.001 kcal/mol was used. Unique low-energy geometries generated were exported and minimized further using a modified CHARMM force field in Catalyst (version 4.11, Accelrys. Inc., San Diego, CA)25 running on a Dell Precision T7400n Mini-Tower, Quad Core Xeon Proc X5450 dual processor computer with operating system, Red Hat Linux 5.0. A maximum number of 250 conformations for each compound was generated using the CHARMM force field parameters and Poling algorithm37 with the “best conformation generation” option and energy threshold of 20 kcal/mol above the lowest-energy conformer to ensure maximum coverage of the conformational space.38,

39

All

other parameters were set to default settings. While generating the hypothesis using HypoGen module from Catalyst, the minimum and maximum count of features for hydrogen bond acceptor 6


Page 7 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(HBA), hydrogen bond donor (HBD), hydrophobic (HY), and aromatic ring (AR) features were set to 0 and 5 per feature type. The default uncertainty value of 3 was used for the compound activity, representing the ratio of uncertainty range of the measured biological activity against the actual activity for each compound. An initial analysis of the function mapping tools revealed HBA, HBD, HY, and AR features effectively mapped all critical features of the training set compounds.

Hence, the four features were selected to form the basis for the hypothesis

generation process. Pharmacophore Model Validation. Hypothesis models were used to predict the activity of the test set of known CXCR2 inhibitors (Supplementary Table 2). The geometric fit value for each of the test set compounds was calculated in order to predict activity, which is a linear function of the fit value. The fit value is a quantitative term used to measure the number of features (HBA, HBD, HY, AR) mapped onto a compound and the distance that separates the feature on the molecule from the centroid of the hypothesis.40 Fit values indicate how well features of a pharmacophore map the select functional groups of a compound. The Güner-Henry (GH) scoring method was used to calculate the “goodness-of-hit”.40

To assess the quality of the

pharmacophore models, a database of active and inactive inhibitors from literature were evaluated with the equation below: =

( + ) − × − × −

where, D is the total compounds in the database, A is the number of actives, Ha is the total number of actives in the hit list, Ht is the total number of compounds in the hit list. For this assessment, a database of known active and inactive inhibitors (previously published) were used to assess the pharmacophore.30-32, 34, 35, 41-55 GH scores were calculated for the top 10 hypotheses. 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 70

The GH score values range from 0 (null model) to 1(ideal model). A GH score value of greater than 0.7 indicates a good predictive model.40 Database Search. The best pharmacophore hypothesis (Hypo 1) was used as a search query to retrieve compounds with novel structural scaffolds and desired chemical features from a multiconformer Catalyst-formatted database consisting of 5 million commercially available compounds from Asinex (Moscow, Russia), Vitas-M Laboratory (Apeldoorn, Netherlands), and Enamine (Kiev, Ukraine). The fast flexible search method in Catalyst was used to search the database. Molecular Docking. Molecular docking studies were performed using GOLD (Genetic Optimization for Ligand Docking) software package, version 5.2 in the allosteric site of CXCR2 homology model obtained from GPCRR DB.56 The C-terminal loop consisting amino acid 315 to 324 was refined using ModLoop.57 This allowed automated refinement of protein loop, to make allosteric site accessible to solvent and ligand. The quality of the refined model was verified using Ramachandran plot within PROCHECK (http://services.mbi.ucla.edu/SAVES/). A total of 93.5% residues were predicted to be in the most favored region, a 6.5% of residues were in the allowed regions, while none of the residues were in disallowed regions. Prior to docking, 10 different conformations were generated for each ligand using Omega (OpenEye Scientific, Santa Fe, NM), a systematic, knowledge-based conformer generator.58 GOLD uses a genetic algorithm to explore the conformational space of a compound inside the binding site of a protein.26,

27

Docking studies were performed using the standard default settings with 100 genetic algorithm (GA) runs on each molecule and by allowing side chain flexibility of important allosteric site residues within their rotamer library.

8


Page 9 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Generation of Shape-based Pharmacophore. Shape-based pharmacophore was generated using ROCS v3.2.0.4 (OpenEye Scientific).28, 29 ROCS performs shape-based overlays of conformers from a candidate database to a query molecule with one or more conformations. Here, we have used combination of shape overlap and color overlap by aligning groups with similar properties in the color force field. The score, referred to as Tanimoto Combo was calculated as the sum of shape and color score.29 The maximum value of the Combo score is 2 and the minimum value is 0.

Cell Culturing. Colon cancer cell lines HCT116 p53+/+ and HCT116 p53−/− were kindly provided by Dr. Bert Vogelstein (Johns Hopkins Medical Institutions, Baltimore, MD). NSCLC cell lines NCI-H1299, NCI-H460, A549 were purchased from the NCI (Bethesda, MD). HCT116p53+/+, HCT116 p53−/−, H1299, H460, and A549 were grown in RPMI-1640 (Invitrogen, Carlsbad, CA) supplemented with 10% FBS at 37 °C in a humidified atmosphere of 5% CO2. CXCR2-bla and CXCR4-bla U2OS Tango cells were purchased from Invitrogen (Carlsbad, CA) and grown in McCoy5A supplemented with 10% dialyzed FBS, zeocin (200µg/mL), hygromycin (50 µg/mL), geneticin (100 µg/mL), 1 mM sodium pyruvate, 0.1 mM nonessential amino acids (NEAA), and 25 mM HEPES. 293T cells over-expressing GFP-tagged CXCR2 (293T-CXCR2-GFP) were stably generated by Dr. Daryl Davies (University of Southern California, School of Pharmacy, Los Angeles, CA) using a lentiviral system. 293TCXCR2-GFP cells were cultured in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% FBS and puromycin (2 µg/mL). All cells were grown at 37 °C in a humidified atmosphere of 5% CO2. All cell lines used were maintained in culture under 35 passages and tested regularly for mycoplasma contamination using Plasmo TestTM (InvivoGen, San Diego, CA). 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 70

Compounds and Reagents. Compounds were purchased from Asinex (Moscow, Russia) and Enamine (Monmouth, NJ). Compounds were prepared in DMSO at 10 mM stock solution and stored at -20 °C. SB225002 and SB265610 were purchased from Tocris Bioscience (Ellisville, MI). Human CXCL8 cDNA was inserted into the expression vector, pET32a at EcoR1 and XhoI restriction site to generate His-tagged CXCL8. The sequence of the clone was confirmed by DNA sequencing. The recombinant CXCL8 was expressed in BL21-Gold (DE3) pLysS strain of E. Coli (Stratagene, La Jolla, CA) and purified using previously reported protocol.59

CXCR2 and CXCR4 Tango Assay. CXCR2-bla U2OS cells stably over-express CXCR2 linked to a TEV protease site and a GAL4-VP16 transcription factor. These cells also stably express a β-arrestin-2/TEV protease fusion protein and a β-lactamase reporter gene. Upon CXCR2 activation in the presence of CXCL8, the β-arrestin-2/TEV protease fusion protein is recruited to the receptor and cleaves the peptide linker that links CXCR2 to the GAL4-VP16 transcription factor.

Once cleaved, GAL4-VP16 may enter the nucleus and promote the

transcription of the β-lactamase gene. β-lactamase activity is detected using a FRET-based fluorescence assay with CCF4-AM (Invitrogen, Carlsbald, CA). CCF4-AM is a β-lactamase FRET substrate and is cleaved in the presence of β-lactamase. The uncleaved substrate excites at 409 nm and emits at 535 nm, whereas the cleaved substrate excites at 409 nm and emits at 460 nm.

Thus, CXCR2 activity is directly correlated with the amount of cleaved β-lactamase

substrate. In each assay, CXCR2 or CXCR4-bla U2OS cells were seeded (11,000/well) in 384well tissue culture plates for 24 h in DMEM supplemented with 1% dialyzed FBS. Cells were pretreated with various concentrations of inhibitors for 30 min prior to the addition of CXCL8 10


Page 11 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(20 nM) or SDF-1 (30 nM) and incubated for 5 h at 37 °C, unless otherwise noted in the figure legends. β-lactamase substrate was loaded according to the manufacturer’s protocol for 2 h and plates were read on Envision microplate reader (Perkin Elmer) at 409 nm excitation and 460/540 nm emissions. Ratio was calculated for each sample well: cleaved (409/460)/uncleaved (409/540). Percent inhibition was calculated using the following formula: % Inhibition = [1-((Compound treated-unstimulated control)/(CXCL8/SDF stimulatedunstimulated control))] x 100

Cell Proliferation. Cell Proliferation was assessed by MTT assay as previously described.60 In brief, cells were seeded in 96-well microtiter plates (3,000-5,000 cells/well) and allowed to attach overnight. Cells were continuously treated with compounds for 72 h. At the end of treatments, cells were incubated with MTT solution (at a final concentration of 0.5 mg/mL) for 3 h at 37 °C. Supernatant was removed and cells were dissolved with 150 µL of DMSO. Absorbance was read at 570 nm on a microplate reader (Molecular Devices, Sunnyvale, CA). All experiments were performed in triplicates. IC50 was determined for each compound from a plot of log (drug concentration) versus percentage of cells killed.

The data reported with

standard deviations is a representative of at least 3 independent experiments.

CXCR2 In-cell Western assay. CXCR2-bla U2OS cells were seeded at 11,000 cells/well in 384-well tissue culture plates in 40 µL of assay media (DMEM supplemented with 1% dialyzed FBS) and incubated at 37°C. The next morning, assay media was removed and cells were pretreated with 1X concentrations of CX compounds (36 µL total volume) for 30 min. 4 µL of 11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 70

10X (10 fold concentrated) CXCL8 was added to each well (final concentration 50 nM). Cells were incubated with CXCL8 for 1-5 h at 37°C. Compounds and CXCL8 was removed and cells were fixed with 4% formaldehyde (25 µL/well) for 20-30 min at room temperature. Formaldehyde was removed and wells were washed with 1X PBS and stored at 4°C. The next day, each well was blocked in blocking buffer (Licor, NE) for 2 h at room temperature and incubated with CXCR2 antibodies (Santa Cruz Biotechnology, 1:1000 dilution in blocking buffer) for 2 h at room temperature or overnight at 4°C. Cells were then washed with 1X PBS six times and incubated with mouse IRDye 680RD secondary antibodies (25 µL/well, 1:1000 dilution in blocking buffer) for 1 h in the dark. Wells were washed with 1X PBS six times to remove unbound antibodies. To remove excess liquid, plates were spun upside down on top of paper towels at 1200 RPM for 5 min.

Plates were completely dried before imaging and

quantification on the Licor Odyssey bioimager. To detect total CXCR2 expression (receptor degradation), cells were blocked in blocking buffer with 0.3% Triton X100 for 2 h. Primary and secondary antibodies were also diluted in 0.3% Triton X100 in blocking buffer.

Percent

inhibition of receptor internalization and degradation was calculated using the following formula: % Inhibition = [(Compound treated - CXCL8 stimulated)/(unstimulated control – CXCL8 stimulation)] x 100

Neutrophil Chemotaxis. Blood was collected from healthy human volunteers into EDTAsprayed tubes (Greiner Bio-One, Monroe, NC). Human neutrophils were isolated using OneStep Polymorph separation media (Accurate Chemical and Scientific, Westbury, NY) with centrifugation at 500 g for 35 min according to manufacturer’s recommendations. RBC Lysis 12


Page 13 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Buffer (IBI Scientific, Peosta, IA) was used to remove residual erythrocytes. The final pellet of neutrophils was resuspended in HBSS without Ca2+ or Mg2+ (Lonza, Walkersville, MD) supplemented with 1% BSA (Amresco, Solon, OH) and kept on ice. Neutrophils were quantified using Turk’s Blood Diluting Fluid (Ricca Chemical, Arlington, TX) and a hemocytometer. Trypan blue exclusion was used to determine neutrophil viability. All preparations consisted of approximately 95% viable neutrophils. All experiments were conducted with the approval of the University of Southern California Institutional Review Board. Isolated neutrophils were incubated in triplicate with CX4152 over four logs of concentration or vehicle (1% DMSO) for 1 h. Neutrophils were then placed in the top wells of a 96-well MultiScreen-MIC plate with 3 µm pore size (Millipore, Billerica, MA) containing 50 nM of CXCL8 in the bottom wells. Neutrophils were allowed to migrate for 2-4 h at 37°C and 5% CO2. In some experiments, the migrated neutrophils in the bottom wells were stained with Turk’s Blood Diluting Fluid and manually counted with a hemocytometer. In other experiments, the migrated neutrophils were labeled with CyQuant NF dye (Invitrogen, Eugene, OR), and fluorescence intensity was measured using a spectrofluorometer with excitation at 485 nm and emission at 530 nm (Perkin Elmer). Neutrophil count was calculated using a standard curve. Percentage of maximal chemotaxis was calculated using the following formula: % maximal chemotaxis = [(# treated cells migrated with CXCL8 - # untreated cells migrated without CXCL8) / (# untreated cells migrated with CXCL8 - # untreated cells migrated without CXCL8)] x 100 Data was plotted using a sigmoidal Emax model to estimate IC50, the concentration associated with a 50% inhibition of the maximal chemotactic response (IC50).

13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 70

Calcium Flux. 293T-CXCR2-GFP cells were seeded at 30,000 cells/well in 384-well black plates with clear bottom plates in growth medium. The next day, cells were loaded with Fluo-4 NW and probenecid according to the manufacturer’s protocol (Invitrogen). Briefly, dye mixture included in kit was diluted with 10 mL assay buffer and 100 µL of probenecid (25 mM final concentration). Growth media was replaced with 25 µL of the dye mixture and incubated for 30 min at 37°C for 30 min and at room temperature for 30 min (protected from light). To detect calcium flux induced by CX compounds, SB265610, or SB225002, 5X of compounds were added to each well and fluorescence signal was detected immediately on the Envision plate reader with 490 nm excitation and 535 nm emission filters.

To detect CXCL8-stimulated

calcium flux, 5-fold CX compounds and 5-fold CXCL8 (5 µL) was added to each well and fluorescence signal was detected immediately on the Envision plate reader with 490 nm excitation and 535 nm emission filters. Fluorescence signal was normalized to baseline signal prior to any stimulation.

LPS-induced Lung Inflammation. Eight- to ten-week-old male BALB/c mice were obtained from Charles River Laboratories and were maintained in the animal facilities located at Zilkha Neurogenetic Institute (University of Southern California, Los Angeles, CA). A total of 12 mice with three in each treatment group (LPS alone and CX4152 at 50 mg/kg, 5 mg/kg, and 0.5 mg/kg) were used for pharmacodynamic evaluation of LPS-induced airway inflammation. Animals had access to food and water and were kept in controlled laboratory conditions with average temperature of 22°C and a light–dark cycle of 12 h. A single 0.25 mL injection containing CX4152 diluted in 5% DMSO in sesame oil or vehicle only (5% DMSO in sesame oil) was administered subcutaneously 30 min prior to LPS insufflation. All compounds were 14


Page 15 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


aseptically prepared. Acute lung inflammation was induced by intranasal LPS insufflation (1 µg/20 g body weight) under light sedation (100 µL of a ketamine (80 mg/kg) and xylazine (10 mg/kg) solution). Airway leukocytes were collected by bronchoscopy at 24 h. Bronchoalveolar lavage was performed after lethal intraperitoneal administration of Euthasol (75 mg/kg). The trachea was exposed through a midline incision with subsequent insertion of a 26G tracheal tube. The lungs were washed 3 times with 3 different aliquots of PBS (1 mL, 0.8 mL, 0.8 mL) with recovery via syringe aspiration of 86%. BAL samples were placed on ice and centrifuged at 360 g and 4°C for 10 min. The cell pellet was used to evaluate the number of infiltrating leukocytes. Total and differential airway cell counts were determined for each BALF sample using a hemocytometer and Wright stain to examine nuclear morphology under the microscope. All animal work was conducted with the approval of the University of Southern California Institutional Animal Care and Use Committee (Protocol #11675). All studies involving animals are reported in accordance with the ARRIVE guidelines.61, 62

RESULTS Development of a Robust CXCR2 Inhibitor-based Pharmacophore Model. We chose three classes of previously described CXCR2 inhibitors (squaramides, guanidines, and diarylureas; Figure 1A) to develop pharmacophore models to identify novel compounds sharing similar chemical properties arranged in the same spatial orientation from a database of 5 million compounds. A pharmacophore is defined as a set of structural features in a drug molecule that is recognized at a receptor site and is responsible for its biological activity.63 Most common pharmacophore features of a drug are hydrogen bond donors and acceptors, hydrophobic regions, and cation/anions. Using a training set of 20 known CXCR2 inhibitors (Supplementary Table 1), 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 70

we have generated 10 quantitative pharmacophore models (Table 1) with the Catalyst software. All the pharmacophore hypotheses are comprised of four chemical features: hydrogen bond acceptor (HBA), hydrogen bond donor (HBD), hydrophobic (HY), and aromatic ring (AR). We have clustered all hypotheses into different clusters (Supplementary Table 3) using hierarchical complete linkage method. From the cluster analysis it is clear that although all pharmacophores contain four similar features, their arrangements are quite different. For instances, Hypo-1 and Hypo-2 both are in the same cluster up to the number of clusters 6, but the HBD feature of Hypo2 is different from that of Hypo-1 in the arrangement. In fact, all pharmacophores are different in terms of feature arrangement. The best pharmacophore hypothesis, Hypo-1 (Figure 1B-C), was selected for further investigation on the basis of its statistical parameters (Table 1).

Figure 1. Pharmacophore modeling of Hypo-1. (A) The structural classes of CXCR2 inhibitors chosen for pharmacophore generation are squaramides, guanidines, and diarylureas. (B) The best hypothesis model Hypo-1 was generated using the Hypo-Gen (Catalyst, Accelrys). 16


Page 17 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Pharmacophore features are shown in magenta for hydrogen-bond donor (HBD), cyan for hydrophobic group (HY), orange for aromatic group (AR), and green for hydrogen-bond acceptor (HBA). (C) Distances between pharmacophore features are given in angstroms. Representatives of potent training set compounds chosen from three different classes of CXCR2 inhibitors are also mapped on Hypo-1: (D) squaramide (1), (E) guanidine (7) and (F) diarylurea (13) are mapped onto the HypoGen model.

Table 1. Properties from pharmacophore hypothesis generation using compounds from the training set Hypothesisa

Total cost

Error cost

RMS

Correlation (R)

Test set correlation (R)

Featureb

Hypo-1

89.43

74.13

0.83

0.93

0.79

HA, HD, HY, AR

Hypo-2

95.20

80.37

1.14

0.86

0.70

HA, HD, HY, AR

Hypo-3

95.83

80.84

1.16

0.86

0.69

HA, HD, HY, AR

Hypo-4

96.04

81.19

1.18

0.86

0.65

HA, HD, HY, AR

Hypo-5

96.81

81.62

1.20

0.85

0.76

HA, HD, HY, AR

Hypo-6

97.12

82.05

1.22

0.85

0.68

HA, HD, HY, AR

Hypo-7

97.25

82.40

1.23

0.84

0.64

HA, HD, HY, AR

Hypo-8

98.29

83.58

1.28

0.83

0.71

HA, HD, HY, AR

Hypo-9

98.44

83.25

1.26

0.83

0.72

HA, HD, HY, AR

Hypo-10

98.58

83.51

1.27

0.83

0.72

HA, HD, HY, AR

a

Null cost =121.34, Fixed Cost = 81.99, Configuration cost = 13.59. All costs are in units of bits. All pharmacophores consist of four features including one each of hydrogen-bond acceptor (HA), hydrogen-bond donor (HD), hydrophobic (HY) and aromatic ring (AR).

b

The value of the total cost of each hypothesis was much closer to the fixed cost value, which is expected for a good hypothesis. The differences between the null hypothesis cost and the fixed cost, and the null hypothesis cost and the total cost of the best hypothesis (Hypo-1) were 39.35 and 31.91 bits, respectively. The entropy (configuration cost) value of 13.59 for the hypothesis is also within the allowed range of 17.00.64 Experimental and estimated activity data from both training and test set compounds are plotted as shown in Figure 2, and reported in Supplementary Tables 1 and 2, respectively. Hypo-1 has excellent predictive ability for the test set compounds

17



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 70

(correlation coefficient, R = 0.79). Most of the test compounds are predicted with an error value less than 5. For further validation of our pharmacophore model, a database of 252 molecules consisting 153 actives (IC50≤100 nM) and 99 inactives (IC50≥1 µM) that were previously reported in the literature was screened with Hypo-1.30-32,

34, 35, 41-43

We did not consider

moderately active compounds (100 nM to 1 µM) in the validation database to make clear distinguish between actives and inactives. GH score was calculated to show robustness of all ten pharmacophore models (Supplementary Table 4). Hypo-1 successfully retrieved 80 % of actives from the database (Table 2). A GH score of 0.80 indicate significantly good quality model.40

Figure 2. Graph showing the correlation between experimental and predicted activities of 20 training (black) and 50 test set (red) compounds.

18


Page 19 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Hypo-1 validation using a database of previously published CXCR2 Inhibitors Parameters

Value

Total compounds in database (D)

252

Total number of actives in database (A)

153

Total hits (Ht)

132

Active hits (Ha)

122

% Yield of actives

92.4

% Ratio of actives in the hit list

79.7

Enrichment factor (E)

1.5

False negatives

31

False positives

10

GH score (goodness of hit )

0.8

Hypo-1 was mapped onto the three highly active and structurally diverse classes of CXCR2 inhibitors (squaramides, guanidines, and diarylureas) with an excellent fit value (>8) (Figure 1DF). Aromatic and hydrophobic features are mapped onto two phenyl groups of the active training set compounds, while hydrogen bond donor and acceptor are mapped onto one of the amine groups and sulfonyl oxygen or cyano group, respectively. Although, ortho hydroxyl group in one of the phenyl rings is most common in the training set, our best model did not identify that feature except in a few cases (OH group of compound 13 is partially mapped onto the hydrogen bond donor feature, Figure 1F). There are several other pharmacophores, such as Hypo-2, Hypo3, Hypo-4 that also mapped onto that hydroxyl group, but they were not the best pharmacophore in terms of pharmacophore parameters and external validation (Table 1 and Supplementary Table 4). Hypo-1 was used to screen a database of 5 million commercially available compounds from Asinex, Enamine, and VitasM Lab.65-67 The criteria for virtual screening were set to a) fit value >5 and b) compounds should map all features of the pharmacophore. This yielded 350 19



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 70

compounds with a fit value >5. These compounds were chosen for cell-based screening in the CXCR2 Tango assay. Identification of CX Compounds in CXCR2 Screening. Using the CXCR2 Tango assay, we initially screened 144 compounds from 350 pharmacophore hits based on structural diversity as well as commercial availability. 13 compounds that belong to a class of sulfonamides were identified as hits from our initial screen. The best initial hits identified from pharmacophore screening are represented in Table 3 and mapped onto the Hypo-1 pharmacophore model in Figure 3. Our most active hit compound, CX25 (IC50=360 ± 100 nM) was chosen for further lead optimization. These compounds were also counter screened in the CXCR4 Tango assay. Interestingly, most of the active CX compounds also inhibited CXCR4 activity in the Tango assay. For example, CX25 inhibited CXCR4 with an IC50 of 0.59 ± 0.1 µM.

Figure 3. Pharmacophore mapping of CX compounds and CXCR2 inhibitors. (A) CX25, (B) CX86, (C) CX815 and CXCR2 inhibitors, (D) SB656933 and (E) SCH527123 are mapped onto the Hypo-1 model with indicated fit values. 20


Page 21 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


Table 3. CXCR2 inhibition and cancer cell proliferation results of active CX compounds selected from Hypo-1 screening Compound Code

Structure

Fit

CXCR2 Inhibition CXCR4 Inhibition (IC50, µM) (IC50, µM)

MTT (IC50, µM) H1299

H460

HCT116 p53+/+

HCT116 p53-/-

SB225002

6.69

0.19±0.24

>20

3.6±4.9

4.1±2.4

4.6±3.9

5.4±4.1

SB265610

6.72

0.28±0.09

>20

>50

>50

>50

>50

SB656933

7.4

NT

NT

NT

NT

NT

NT

SCH527123

6.5

NT

NT

NT

NT

NT

NT

CX25

6.4

0.36±0.1

0.59±0.1

8.0±6.5

3.9±1.5

3.9±1.9

6.2±5.2

CX86

5.81

1.1±1.4

3.4±3.3

7.3±4.8

6.1±2.3

3.8±0.5

5.5±2.2

21



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

CX815

6.15

0.4±0.1

>50

Page 22 of 70

3.5±0.9

4.1±0.2

5.0±3.7

NT

CXCR2 Inhibition was determined using the CXCR2 Tango assay. Cell proliferation was determined using MTT assay. NT = not tested. Fit = fit value for Hypo-1. IC50 is the mean ± SD of two independent experiments in the CXCR2/4 Tango assay.

22


Page 23 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Lead Optimization and SAR Analysis. Given that some of the CXCR2 inhibitors used to derive Hypo-1 were later shown to act allosterically, we decided to dock CX25 onto the allosteric site of the CXCR2 homology model.56, clinical candidate, SCH527123.

68

We validated our docking model using a

The predicted binding mode of the inhibitor exhibited

interaction with important residues, Thr83, Asp84, Ala249, Tyr314, Gln319 and Lys320, which are in agreement with mutagenesis studies (Figure 4).69 From our docking studies, CX25 was predicted to bind in the allosteric site, where it forms hydrogen bond interactions with Asp84 and Lys320 (Figure 5A). The benzothiadiazole ring system of CX25 formed pi-pi interaction with Tyr314, whereas the quinoline ring system formed hydrophobic interactions with Thr83, Ala249 and Gln319 (Figure 5A). Both docked conformation and pharmacophore alignment of CX25 are quite similar. Three out of four Hypo-1 features were reproduced in the docking of CX25. We observed a hydrogen bond interaction between sulfonyl oxygen and Lys320 in the docking model. Contribution of this interaction to the activity of compounds needs further investigation. Although the sulfur atom in the thiadiazole ring was recognized by Hypo-1 as a hydrogen bond acceptor, docking of CX25 did not reveal any interaction with the S. A shape-based pharmacophore (Figure 5B) was generated based on the docking pose of CX25 and its interactions with the binding site residues using ROCS.28, 29 The pharmacophore is comprised of four rings, one anion, and one hydrogen bond interaction, which mimics the docking interactions of CX25. On the basis of the docking and Hypogen pharmacophore models we concluded that sulfonamide NH acts as hydrogen bond donor, but ROCS does not considers this as a hydrogen bond donor, instead ROCS considers this an anion. ROCS pharmacophore not only retrieves compounds with similar features, but also compounds with 3D similarity. We further identified and tested 103 additional sulfonamides from the shape-based pharmacophore search containing 23



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 70

an NH group, a feature recognized by ROCS pharmacophore as an anion. Our docking model accurately recognizes the NH functionality as a hydrogen bond donor.

Figure 4. Docking of SCH527123 onto the allosteric site of CXCR2. Left panel shows the 7 transmembrane of CXR2 homology model (blue colored) with SCH527123 (brown colored) docked onto the allosteric site located inside the receptor. Right panel is the zoomed docking pose of the inhibitor showing interactions with important allosteric site residues (cyan) within the receptor surface cavity (grey) after a 90° rotation in the horizontal plane. Green dotted lines indicate hydrogen bonding interactions.

24


Page 25 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Lead optimization is performed using docking and generating shape-based pharmacophore of CX25. (A) CX25 is docked onto the allosteric site of CXCR2 (grey colored receptor surface). Asp84 and Lys320 formed hydrogen bond interactions with CX25 (purple), while Tyr314 is involved in pi-pi interactions. Several other residues, such as Thr83, Ala249, and Gln319 formed hydrophobic interactions with CX25. (B) A six-feature, shape-based ROCS pharmacophore was generated comprising of four rings, one anion and one hydrogen bond acceptor (HBA) using the docking pose conformation of CX25. (C) The ROCS pharmacophore is mapped onto CX4152 with ROCS Tanimoto Combo score of 1.06. (D) CX4152 (pink) is docked into the allosteric site of CXCR2. It forms two hydrogen bonds with Ser81 and Lys320 and hydrophobic and pi-pi interactions with other important residues. Testing these compounds in the same assay described above resulted in 33 compounds with IC50 values 20

>20

>20

CX279

6.03

1.396

2.2

3.0±1.4

2.1±1.8

2.2±1.8

1.6

CX684

5.94

1.387

1.4±1.1

1.0±0.6

1.2±0.7

0.9±0.7

2.5

CX061

6.18

1.33

2.5±1.7

2.5±1.9

2.3±1.9

1.5±1.1

3.3

CX297

5.88

1.388

2.4±2.3

3.6±1.2

3.0±0.7

4.1±1.3

NT

28


Page 29 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


CX296

5.82

1.459

5.6±3.2

20.7±17.2

45.0±7.3

70.8

NT

CX388

6.13

1.149

1.6

5.0

3.2

5.0

14

CX310

5.86

1.279

1.6±1.2

NT

NT

NT

8.2

CX324

5.86

1.075

3.1±2.7

4.8±3.2

3.0±1.4

2.6±1.4

NT

CX128

5.9

0.882

2.4±2.3

10

>10

NT

29



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 30 of 70

CX826

6.16

1.349

1.7±2.0

7.1

1.4±1.0

0.9±0.6

5.0

CX073

6.06

1.255

7.0±6.1

>20

20.0

17.8

NT

CX587

6.21

1.296

5.0

NT

8.5±0

>10

NT

CXCR2 Inhibition was determined using the CXCR2 Tango assay. Cell proliferation was determined using MTT assay. NT = not tested. Fit = fit value for Hypo-1. ROCS score = Tanimoto combo score for ROCS pharmacophore. IC50 is the mean ± SD of at least two independent experiments in the CXCR2/4 Tango assay.

30


Page 31 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


Table 5. CXCR2 inhibition and cancer cell proliferation results of CX compounds with Scaffold C selected from ROCS pharmacophore of CX25 Compound Code

Structure

Fit

ROCS score

MTT (IC50, µM)

CXCR2 Inhibition (IC50, M)

CXCR4 Inhibition (IC50, µM)

H1299

H460

CX309

6.37

1.105

16.5±1.1

8.9

6.2±5.6

>20

CX640

6.33

1.142

8.4±4.6

NT

>10

>10

CX078

7.18

1.085

8.0±6.8

10.0

NT

NT

CX459

6.46

0.743

3.6±6.9

11.7

8.3±8.2

17.8

31



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 32 of 70

CX910

7.05

0.960

>20

NT

NT

NT

CX588

7.74

0.757

3.6

NT

NT

NT

CX886

7.76

0.748

7.9

NT

3.6±0.6

15.8

CX887

6.57

0.974

15.4±1.1

8.9

NT

NT

32


Page 33 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


CX410

6.54

0.923

17.8

NT

5.0±0

>10

CX396

6.34

0.849

6.5±8.7

7.6

NT

NT

CX404

6.34

0.668

11.1±6.8

20

NT

NT

CX474

6.42

1.194

16.3±3.4

>20

20.0

>20

33



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 34 of 70

CX489

7.12

1.178

>20

NT

NT

NT

CX4152

6.21

1.028

7.6±6.2

64.7±18.9

31.6

20.0

CXCR2 Inhibition was determined using the CXCR2 Tango assay. Cell proliferation was determined using MTT assay. NT = not tested. Fit = fit value for Hypo-1. ROCS score = Tanimoto combo score for ROCS pharmacophore.

34


Page 35 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


All compounds with scaffold B were inactive at 10 µM (Supplementary Table 5).

For

example, compound CX624 was inactive at 10 µM, whereas compound CX061 (scaffold A) had an IC50 value of 2.5 µM. These data suggest that the NH group (hydrogen bond donor) on the quinoline is essential for activity as it is a significant pharmacophoric feature. Compounds with naphthalene instead of quinoline (scaffold C, Table 5) were slightly less active than compounds with scaffold A, suggesting the importance of the nitrogen in the quinoline structure. Although there is no interaction between quinoline nitrogen and the active site residues in our docked model, the importance of quinoline may be attributed to its physicochemical properties affecting cell permeability and solubility, showing slightly better activity than its naphthalene analogs.

CX4152 is Selective for CXCR2 over CXCR4. CX4152 (Table 5, Figure 7A) showed better selectivity for CXCR2 (IC50=7.6±6.2 µM) over CXCR4 (IC50=64.7±18.9 µM) in the Tango assays (Figure 7B) and was selected for further mechanistic studies. Figure 7C shows that order of addition of CX4152 and CXCL8 affects the potency of CX4152 in the CXCR2 Tango assay. CX4152 showed greater potency when CXCR2-bla U2OS cells were pretreated with CX4152 for 30 min prior to CXCL8 stimulation than when cells were co-treated with CX4152 and CXCL8 at the same time. CX4152 is mapped onto the ROCS pharmacophore with Tanimoto Combo score of 1.06 (Figure 5C) and also fits well into the Hypo-1 pharmacophore model with a fit value of 6.21 (Figure 7D). When CX4152 was docked in the allosteric site of CXCR2 it forms hydrogen bond interactions with Ser81 and Lys320, hydrophobic interactions with Thr83, Ala249 and Gln319, and pi-pi interactions with Tyr314 (Figure 5D). The docking pose of CX4152 reproduced four out of four Hypo-1 features as well as three out of four ROCS pharmacophore features. We also performed similarity searches with CX4152 and found additional compounds 35



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 70

that exhibit similar chemical features and shape. However, these analogs were not more potent than CX4152 and showed less CXCR2 selectivity (Table 6).

Figure 7. CX4152 selectively inhibits CXCR2. (A) Chemical structure of CX4152. (B) CX4152 is mapped onto Hypo-1 with a fit value of 6.21. (C) Dose response curves of CX4152 in the CXCR2 and CXCR4 Tango assay. Cells were pretreated with various doses of CX4152 for 30 min prior to CXCL8 (30 nM) stimulation for 5 h. (D) CX4152 inhibits CXCR2 in the Tango assay in a time-dependent manner. Co-treatment of CX4152 and CXCL8 reduced the potency of CX4152 in the CXCR2 Tango assay compared to cells pretreated with CX4152 for 30 min prior to CXCL8 stimulation. Data represents mean ± SD of a representative experiment performed in duplicates. * indicates p-value < 0.05.

36


Page 37 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


Table 6. Inhibition of CXCR2 and CXCR4 activation and cytotoxicity of CX4152 analogues Compound Code

Compound

CXCR2 ROCS score (IC50, µM)a

CXCR4 (IC50, µM)a

MTT (IC50, µM)b 293TCXCR2

293T

H460

H1299

A549

CX8794

1.029

40±13.2

4

2

2.5

3.1

3.5

3.5

CX8793

1.012

28.3±22.5

1

100

40

70

CX4138

0.955

27.5±24.7

18

36

33

22

29

39

CX0965

0.833

32±39.6

1

1.1

3

3.5

4

4

CX0572

1.12

52.3±44.5

30

NT

NT

NT

NT

NT

38


Page 39 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


CX0570

1.05

12.4±11.5

25

4.6

4.5

4.7

5

7

CX0562

1.053

18.7±22.8

6

NT

NT

NT

NT

NT

CX8787

1.028

6.3±1.1

30

20

55

10

55

60

CX0994

1.022

>100

NT

NT

NT

NT

NT

NT

39



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 40 of 70

CX8226

1.05

>100

NT

NT

NT

NT

NT

NT

CX8732

0.867

>100

NT

NT

NT

NT

NT

NT

CX7769

1.025

>100

NT

NT

NT

NT

NT

NT

CX6232

0.781

>100

NT

NT

NT

NT

NT

NT

40


Page 41 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48


CX4093

1.047

>100

NT

NT

NT

NT

NT

NT

a

IC50 is the mean ± SD of three independent experiments in the CXCR2/4 Tango assay. bCells were treated with compounds for 72 h and cell proliferation measured using the MTT assay. cNT is not tested. ROCS score = Tanimoto combo score for ROCS pharmacophore.

41



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 42 of 70

CX4152 Induces CXCR2 Internalization. To characterize the mechanism of CX4152, we assessed its effects on receptor internalization using in-cell Western assays. CX4152 induced receptor internalization in a dose- and time-dependent manner (Figures 8A and B). We also observed a decrease in total CXCR2 at 5 h treatment (Figures 8C and D). Increased in total CXCR2 receptor was observed at earlier time points. Treatment with CX25 and SB265610 did not induce receptor internalization as observed with CX4152 (Figure 9).

Figure 8. CX4152 induces CXCR2 internalization and turnover. CXCR2 surface and total expression was measured using a CXCR2 receptor in-cell Western assay. CXCR2-bla U2OS cells were treated with CX4152 at indicated concentrations for various time points. (A) CX4152 dose and time dependently reduced CXCR2 surface expression. (B) Quantification of data in panel A. Data represents mean ± SD of a representative of at least three independent experiments performed in duplicates. (C) CX4152 reduced total CXCR2 expression at the 5 h time point. (D) Quantification of data in panel C. Data represents mean ± SD of a representative of at least 3 independent experiments performed in duplicates.


42

Page 43 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 9. CX25 does not induce CXCR2 internalization and turnover. CXCR2-bla U2OS cells were treated with various doses of CX25 and CX4152 for 1 h. Cells were also treated with 10 µM of SB265610 for h. Surface (A) and total (B) expression of CXCR2 was assessed using the CXCR2 receptor in-cell Western assay. Quantification of A and B shows that CX25 and SB265610 treatment did not reduce surface (C) and total (D) CXCR2 expression.

CX4152 Does Not Induce Calcium Mobilization. Since we showed that CX25 induced rapid calcium mobilization in lung and colon cancer cells, we also assessed the effects of CX4152 on calcium mobilization.

CX4152 did not induce rapid calcium mobilization in CXCR2-

overexpressing cells (293T-CXCR2-GFP), whereas CX25 induced peak calcium flux within 1 min stimulation (Figure 10). SB265610 alone did not have any effects on calcium mobilization. Stimulation with 300 nM of CXCL8 rapidly induced transient calcium mobilization. SB265610 inhibited CXCL8-induced calcium flux as indicated by a reduced calcium peak when cells were


43


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 44 of 70

co-stimulated with SB25610 and CXCL8. CX4152 had no effects on CXCL8-induced calcium flux.

Figure 10. Calcium response curves of CX4152, CX25, and SB265610 in CXCR2 overexpressing cells. 293T-CXCR2 cells were stimulated with CX4152 (50 µM), SB265610 (5 µM), CX25 (25 µM), or CXCL8 (300 nM). Cells were also co-stimulated with CX4152 (50 µM) and CXCL8 (300 nM) or SB265610 (5 µM) and CXCL8 (300 nM). CX4152 alone did not induce calcium mobilization and had no effects on CXCL8-mediated calcium mobilization.

CX4152 Inhibits Cell Proliferation. We further assessed the cellular effects of CX4152 on proliferation of CXCR2-expressing cells (293T-CXCR2-GFP and CXCR2-bla U2OS) and its parental cell line (293T, U2OS) as well as CXCR4-expressing (CXCR4-bla U2OS) and NSCLC (A549, H460, and H1299) cell lines. CX4152 inhibited cell proliferation (72 hr treatment) in all the cell lines tested with similar IC50s between CXCR2-overexpressing and parental cell lines, suggesting that the effects on cell proliferation may not be CXCR2-mediated (Table 7).


44

Page 45 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 7. Cytotoxicity of CX4152 in a panel of cancer cell lines Cell Lines

MTT IC50 (µM)

293T

24.5±6.4

293T-CXCR2

21±9.9

CXCR2-bla U2OS

42, >50

CXCR4-bla U2OS

44, >50

U2OS

37.5±9.2

A549

21±9.9

H460

20

H1299

31.6

IC50 is the mean ± SD of 2-3 independent MTT experiments performed in triplicates. Cells were treated with compounds for 72 h and cell proliferation measured using the MTT assay.

CX4152 Inhibits PMN Cell Migration. Lastly, to assess the anti-chemotactic effects of CX4152, we measured human neutrophils chemotaxis toward 50 nM of CXCL8 using a modified Boyden chamber transwell assay. We previously determined 50 nM of CXCL8 was an optimal concentration (data not shown). CX4152 inhibited CXCL8-induced chemotaxis in a concentration-dependent manner with an IC50 of 51 µM (Figure 11A). More importantly, 50 mg/kg of CX4152 was able to significantly inhibit PMN migration (about 2-fold decrease) using a murine model of neutrophilic airway inflammation induced with LPS (Figure 11B).


45


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 46 of 70

Figure 11. CX4152 inhibits CXCL8-mediated cell migration. (A) Concentration-dependent inhibition of human neutrophil chemotaxis. Neutrophils were incubated with CX4152 for 1 h, and placed in the top wells of a chemotaxis plate containing 50 nM of CXCL8 in the bottom wells. Neutrophils were allowed to migrate for 2-4 h at 37°C and 5% CO2. Inhibition of chemotaxis was evaluated based on cell counts relative to control (untreated with CXCL8). CX4152 concentration-dependently inhibited CXCL8-induced chemotaxis with an IC50 of 58 µM (95% CI 44.5-76.3, r2 = 0.93). Results are mean ± SD of 3 experiments. (B) Attenuation of neutrophil recruitment in BALF of mice at 24 h by pretreatment with CX4152 prior to intranasal LPS challenge. CX4152 (50 mg/kg subcutaneously) significantly inhibited neutrophil influx (*, p < 0.05). Results are mean ± SD of 3 animals per treatment group (N=12).

DISCUSSION AND CONCLUSIONS The identification of compounds that modulate CXCR2 function was facilitated by computational studies using a CXCR2 inhibitor-based pharmacophore model, Hypo-1. This model exhibits high predictability power to identify CXCR2 inhibitors with a test set correlation between predicted activity and experimental activity of 0.79 (Table 1 and Figure 2) and a good GH score (0.80). This model discriminates between actives (fit > 5) and inactives in an in-house database of CXCR2 inhibitors, reducing the need to perform large-scale in-vitro screening studies (Table 2).

Virtual screening of a commercially available database of 5 million

compounds with Hypo-1 retrieved 350 compounds with fit values > 5. Thirteen out of 144


46

Page 47 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


compounds selected for in-vitro screening showed IC50’s < 10 µM, a hit rate of 9.7%. Of these, CX25 was the most potent, inhibiting CXCR2 activation with an IC50 of 360 nM, and was selected for further studies to characterize its effects in cancer cell lines (Table 3). Hypo-1 also identified four chemical features (hydrogen bond donor, hydrophobic, hydrogen bond acceptor, and ring aromatic) that are important for CXCR2 inhibition. The location of these chemical features is important for activity, which is clearly demonstrated with scaffolds A and B. Activity was

abolished

when

the

hydrophobic

group

(benzene)

and

the

aromatic

group

(quinoline/naphthalene) between the sulfonamide backbone were swapped. CX compounds (CX25, CX86, CX815, CX122) showing nanomolar activity against CXCR2 in the Tango assay also inhibited cell migration, proliferation, and colony formation in cancer cell lines with similar potencies as SB225002 (Supplementary Figures 1-3). Additionally, we observed that SB225002 arrested cells in the G2/M phase, which is inconsistent with previous studies that showed that CXCR2 knockdown in ovarian cancer cell lines arrested cells in the G0/G1 phase (Supplementary Figure 4).18 However, SB265610, a derivative of SB225002 that is just as potent at inhibiting CXCR2 in the Tango assay as SB225002, did not show cytotoxicity in any cell lines tested and inhibited cell proliferation less than 50% at 50 µM. This is consistent with previous studies showing that CXCR2 knockdown with shRNA in melanoma cells reduced cell proliferation by 40% in low serum.70 CX compounds arrested cell cycle in the S phase in NSCLC cell lines and arrested cells in the G2/M phase in colon cancer cell lines (Supplementary Figure 4). This suggests that the toxicity observed with CX compounds and SB225002 may not be CXCR2-mediated, but rather an off-target effect of these compounds. Indeed, SB225002 was recently shown to exhibit microtubule destabilizing activity by binding to tubulin and inducing mitotic arrest in p53 mutant cancer cell lines.71, 72 It was also reported that SB265610 did not


47


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 48 of 70

induce G2/M cell cycle arrest at concentrations as high as 100 µM. This indicates that the slight difference in chemical structure between SB225002 and SB265610 dramatically alters the offtarget effects of inhibitors belonging to the diarylurea class. This further suggests that the Hypo-1 pharmacophore model may also capture compounds exhibiting off-target effects that are inherent in the CXCR2 inhibitors (training set) used to generate Hypo-1. We propose that the cytotoxicity of CX25 may be due to its effects on calcium release. CX25 induced a rapid burst of intracellular calcium release in a dose dependent manner (5-20 µM) in lung and colon cancer cells (Supplementary Figure 5), whereas SB225002 and SB265610 only slightly induced calcium at higher concentrations (20 µM). Calcium signaling is a key component of apoptotic signaling. Agents that increase intracellular free calcium such as thapsigargin (SERCA inhibitor) and ionomycin (calcium ionophore) have been shown to induce apoptosis.73, 74 A counter screen with another chemokine receptor, CXCR4, showed that CX compounds also inhibited CXCR4 with similar potencies suggesting that these compounds are not very selective for CXCR2 and may target other chemokine receptors as well as similar classes of GPCRs. Additional SAR and optimization of these compounds may improve their selectivity. Thus, we performed additional shape-based ROCS pharmacophore searches and screening. We identified CX4152 to be eight times more selective for CXCR2 over CXCR4 in the Tango assay, with IC50’s of 7.6 ±6.2 and 64.7 ± 18.9 µM, respectively. CXCR2 and CXCR4 belong to the same chemokine family and are activated by different ligands. Every component of the CXCR2 and CXCR4 Tango assay are identical, except for chemokine receptor expression in the Tango cell lines and ligand activation. Thus, we do not anticipate selectivity in CXCR2/4 Tango assays if β-arrestin-2 is the target of inhibition or if the compounds target any component of the assay (i.e.


48

Page 49 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


fluorescence readout, reporter gene expression). As observed with SB265610 and SB225002, CX4152 exerts very different cellular effects than CX25 despite sharing similar chemical backbone structure; further demonstrating that slight chemical modifications dramatically alter off-target effects and perhaps selectivity. CX4152 reduced cell proliferation in a panel of cell lines with high and low CXCR2 expression (Table 7); suggesting that the anti-proliferative effects of CX4152 is most likely CXCR2-independent, perhaps targeting other chemokine receptors or other cellular targets. CX4152 induced receptor internalization within 10 minutes, suggesting that inhibitory effects of CX4152 in the CXCR2 Tango assay are due to β-arrestin-2independent receptor internalization. This phenomenon is most likely not due to receptor downregulation since total CXCR2 expression was not decreased until five hours later. Rather, at earlier time points total CXCR2 appears to increase, perhaps due to the accumulation of intracellular CXCR2. This is further supported by studies that show CX4152 was more potent at inhibiting β-arrestin-2 recruitment in the CXCR2 Tango assay when cells were pretreated with CX4152 prior to CXCL8 stimulation than when cells were treated with CX4152 and CXCL8 at the same time (Figure 7C). These observations suggest that CX4152 may be acting as an agonist, however, it did not activate calcium mobilization (Figure 10).

We postulate that

CX4152 inhibits CXCR2/β-arrestin-2 association by down-regulating surface expression of CXCR2 (Figure 12). Through this mechanism, CX4152 inhibited CXCL8-mediated PMN cell migration and LPS-induced lung inflammation in mice (Figure 11). This mechanism of action has not been described in any previously reported CXCR2 antagonists to date.


49


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 50 of 70

Figure 12. Schematic illustration of CX4152 proposed mechanism of action. CX4152 induces receptor internalization in time- and dose-dependent manner. CXCR2 internalization also leads receptor degradation, perhaps via lysosomes, resulting in reduction of total CXCR2. This subsequently leads to decreased CXCR2/β-arrestin-2 association upon CXCL8 activation in the Tango assay. This is supported by data showing that CX4152 inhibitory activities in the Tango assay is dependent on pre-treatment time.

ASSOCIATED CONTENT Supporting Information Supporting materials and methods, figures and tables are included as supplementary information. Also, structures of compounds external validation database are included as csv file format as supplementary file.

AUTHOR INFORMATION


50

Page 51 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Corresponding Author *Phone: 734-647-2732. E-mail: [email protected]. Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes The authors declare no competing financial interest.

ACKNOWLEDGEMENTS We acknowledge the funding support from the Department of Defense, Lung Cancer Research Program Concept Award, Grant #LC090363. We thank Dr. Young-Kwon Hong (Keck School of Medicine, University of Southern California) for generously providing the CXCL-8 cDNA clone.

ABBREVIATIONS USED CXCR2, CXC receptor-2; CXCR1, CXC receptor-1; CXCR4, CXC receptor-4; CXCL8, interleukin-8; ENA-78, epithelial cell-derived neutrophil-activating protein 78; NAP-2, neutrophil-activating peptide 2; ELR+, glutamate, leucine, arginine motif; COPD, chronic obstructive pulmonary disorder; CF, cystic fibrosis; GH, Güner-Henry; HBD, hydrogen bond donor; HBA, hydrogen bond acceptor; HY, hydrophobic; AR, aromatic ring; NEAA, nonessential amino acids; HEPES, 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid; TEV, tobacco etch virus; FRET, fluorescence resonance energy transfer; SDF-1, stromal cell-derived factor 1; DMEM, Dulbecco’s Modified Eagle Medium; PBS, phosphate buffered saline; FBS, fetal bovine serum; MTT, (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; DNA, deoxyribonucleic acid; RPMI, Roswell Park Memorial Institute; RPM, rotation per minute; EDTA, Ethylenediaminetetraacetic acid; DMSO, dimethyl sulfoxide; BALF, bronchoalveolar lavage fluid; LPS, lipopolysaccharide; GOLD, genetic optimization for ligand docking; ROCS, Rapid Overlay of Chemical Structures


51


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 52 of 70

REFERENCES 1. 2.

3.

4. 5.

6.

7. 8.

9.

10.

11.

12.

13.

14.

Baggiolini, M.; Dewald, B.; Moser, B., Interleukin-8 and Related Chemotactic Cytokines-CXC and CC Chemokines. Adv. Immunol. 1994, 55, 97-179. Xu, L.; Kelvin, D. J.; Ye, G. Q.; Taub, D. D.; Ben-Baruch, A.; Oppenheim, J. J.; Wang, J. M., Modulation of Il-8 Receptor Expression on Purified Human T Lymphocytes Is Associated with Changed Chemotactic Responses to Il-8. J. Leukoc. Biol. 1995, 57, 335342. Chuntharapai, A.; Lee, J.; Hebert, C. A.; Kim, K. J., Monoclonal Antibodies Detect Different Distribution Patterns of Il-8 Receptor a and Il-8 Receptor B on Human Peripheral Blood Leukocytes. J. Immunol. 1994, 153, 5682-5688. Murdoch, C.; Monk, P. N.; Finn, A., CXC Chemokine Receptor Expression on Human Endothelial Cells. Cytokine 1999, 11, 704-712. Horuk, R.; Martin, A. W.; Wang, Z.; Schweitzer, L.; Gerassimides, A.; Guo, H.; Lu, Z.; Hesselgesser, J.; Perez, H. D.; Kim, J.; Parker, J.; Hadley, T. J.; Peiper, S. C., Expression of Chemokine Receptors by Subsets of Neurons in the Central Nervous System. J. Immunol. 1997, 158, 2882-2890. Ahuja, S. K.; Murphy, P. M., The CXC Chemokines Growth-Regulated Oncogene (Gro) Alpha, Grobeta, Grogamma, Neutrophil-Activating Peptide-2, and Epithelial Cell-Derived Neutrophil-Activating Peptide-78 Are Potent Agonists for the Type B, but Not the Type a, Human Interleukin-8 Receptor. J. Biol. Chem. 1996, 271, 20545-20550. Murphy, P. M.; Tiffany, H. L., Cloning of Complementary DNA Encoding a Functional Human Interleukin-8 Receptor. Science 1991, 253, 1280-1283. Coffer, P. J.; Geijsen, N.; M'Rabet, L.; Schweizer, R. C.; Maikoe, T.; Raaijmakers, J. A.; Lammers, J. W.; Koenderman, L., Comparison of the Roles of Mitogen-Activated Protein Kinase Kinase and Phosphatidylinositol 3-Kinase Signal Transduction in Neutrophil Effector Function. Biochem. J. 1998, 329 ( Pt 1), 121-130. Constantin, G.; Majeed, M.; Giagulli, C.; Piccio, L.; Kim, J. Y.; Butcher, E. C.; Laudanna, C., Chemokines Trigger Immediate Beta2 Integrin Affinity and Mobility Changes: Differential Regulation and Roles in Lymphocyte Arrest under Flow. Immunity 2000, 13, 759-769. Addison, C. L.; Daniel, T. O.; Burdick, M. D.; Liu, H.; Ehlert, J. E.; Xue, Y. Y.; Buechi, L.; Walz, A.; Richmond, A.; Strieter, R. M., The CXC Chemokine Receptor 2, CXCR2, Is the Putative Receptor for Elr+ CXC Chemokine-Induced Angiogenic Activity. J. Immunol. 2000, 165, 5269-5277. Nocker, R. E.; Schoonbrood, D. F.; van de Graaf, E. A.; Hack, C. E.; Lutter, R.; Jansen, H. M.; Out, T. A., Interleukin-8 in Airway Inflammation in Patients with Asthma and Chronic Obstructive Pulmonary Disease. Int. Arch. Allergy Immunol. 1996, 109, 183-191. Yamamoto, C.; Yoneda, T.; Yoshikawa, M.; Fu, A.; Tokuyama, T.; Tsukaguchi, K.; Narita, N., Airway Inflammation in Copd Assessed by Sputum Levels of Interleukin-8. Chest 1997, 112, 505-510. Busch-Petersen, J., Small Molecule Antagonists of the CXCR2 and CXCR1 Chemokine Receptors as Therapeutic Agents for the Treatment of Inflammatory Diseases. Curr Top Med Chem 2006, 6, 1345-1352. Widdowson, K. L. Phenyl Urea Interleukin-8 Receptor Antagonists for Treatment of Interleukin-8-Mediated Diseases, and Preparation Thereof. WO199749286, 1997.


52

Page 53 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


15. Zhu, Y. M.; Webster, S. J.; Flower, D.; Woll, P. J., Interleukin-8/CXCL8 Is a Growth Factor for Human Lung Cancer Cells. Br. J. Cancer 2004, 91, 1970-1976. 16. Rubie, C.; Kollmar, O.; Frick, V. O.; Wagner, M.; Brittner, B.; Graber, S.; Schilling, M. K., Differential CXC Receptor Expression in Colorectal Carcinomas. Scand. J. Immunol. 2008, 68, 635-644. 17. Murphy, C.; McGurk, M.; Pettigrew, J.; Santinelli, A.; Mazzucchelli, R.; Johnston, P. G.; Montironi, R.; Waugh, D. J., Nonapical and Cytoplasmic Expression of Interleukin-8, CXCR1, and CXCR2 Correlates with Cell Proliferation and Microvessel Density in Prostate Cancer. Clin. Cancer Res. 2005, 11, 4117-4127. 18. Yang, G.; Rosen, D. G.; Liu, G.; Yang, F.; Guo, X.; Xiao, X.; Xue, F.; Mercado-Uribe, I.; Huang, J.; Lin, S. H.; Mills, G. B.; Liu, J., CXCR2 Promotes Ovarian Cancer Growth through Dysregulated Cell Cycle, Diminished Apoptosis, and Enhanced Angiogenesis. Clin. Cancer Res. 2010, 16, 3875-3886. 19. Luppi, F.; Longo, A. M.; de Boer, W. I.; Rabe, K. F.; Hiemstra, P. S., Interleukin-8 Stimulates Cell Proliferation in Non-Small Cell Lung Cancer through Epidermal Growth Factor Receptor Transactivation. Lung Cancer 2007, 56, 25-33. 20. Ning, Y.; Manegold, P. C.; Hong, Y. K.; Zhang, W.; Pohl, A.; Lurje, G.; Winder, T.; Yang, D.; LaBonte, M. J.; Wilson, P. M.; Ladner, R. D.; Lenz, H. J., Interleukin-8 Is Associated with Proliferation, Migration, Angiogenesis and Chemosensitivity in Vitro and in Vivo in Colon Cancer Cell Line Models. Int. J. Cancer 2011, 128, 2038-2049. 21. Waugh, D. J.; Wilson, C., The Interleukin-8 Pathway in Cancer. Clin. Cancer Res. 2008, 14, 6735-6741. 22. Li, A.; Varney, M. L.; Singh, R. K., Constitutive Expression of Growth Regulated Oncogene (Gro) in Human Colon Carcinoma Cells with Different Metastatic Potential and Its Role in Regulating Their Metastatic Phenotype. Clin. Exp. Metastasis 2004, 21, 571-579. 23. Ogata, H.; Sekikawa, A.; Yamagishi, H.; Ichikawa, K.; Tomita, S.; Imura, J.; Ito, Y.; Fujita, M.; Tsubaki, M.; Kato, H.; Fujimori, T.; Fukui, H., Groalpha Promotes Invasion of Colorectal Cancer Cells. Oncol. Rep. 2010, 24, 1479-1486. 24. Keane, M. P.; Belperio, J. A.; Xue, Y. Y.; Burdick, M. D.; Strieter, R. M., Depletion of CXCR2 Inhibits Tumor Growth and Angiogenesis in a Murine Model of Lung Cancer. J. Immunol. 2004, 172, 2853-2860. 25. Catalyst 4.11 Version; Accelrys Inc, San Diego, CA, 2005. http://www.accelrys.com 26. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R., Development and Validation of a Genetic Algorithm for Flexible Docking. J. Mol. Biol. 1997, 267, 727-748. 27. Jones, G.; Willett, P.; Glen, R. C., Molecular Recognition of Receptor Sites Using a Genetic Algorithm with a Description of Desolvation. J. Mol. Biol. 1995, 245, 43-53. 28. ROCS 3.2.0.4: Openeye Scientific Software, Santa Fe, NM. http://www.eyesopen.com 29. Hawkins, P. C.; Skillman, A. G.; Nicholls, A., Comparison of Shape-Matching and Docking as Virtual Screening Tools. J. Med. Chem. 2007, 50, 74-82. 30. Dwyer, M. P.; Yu, Y.; Chao, J.; Aki, C.; Chao, J.; Biju, P.; Girijavallabhan, V.; Rindgen, D.; Bond, R.; Mayer-Ezel, R.; Jakway, J.; Hipkin, R. W.; Fossetta, J.; Gonsiorek, W.; Bian, H.; Fan, X.; Terminelli, C.; Fine, J.; Lundell, D.; Merritt, J. R.; Rokosz, L. L.; Kaiser, B.; Li, G.; Wang, W.; Stauffer, T.; Ozgur, L.; Baldwin, J.; Taveras, A. G., Discovery of 2-HydroxyN,N-Dimethyl-3-{2-[[(R)-1-(5Methylfuran-2-Yl)Propyl]Amino]-3,4-Dioxocyclobut-1Enylamino}Benzamide (Sch 527123): A Potent, Orally Bioavailable CXCR2/CXCR1 Receptor Antagonist. J. Med. Chem. 2006, 49, 7603-7606.


53


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 54 of 70

31. Jin, Q.; Nie, H.; McCleland, B. W.; Widdowson, K. L.; Palovich, M. R.; Elliott, J. D.; Goodman, R. M.; Burman, M.; Sarau, H. M.; Ward, K. W.; Nord, M.; Orr, B. M.; Gorycki, P. D.; Busch-Petersen, J., Discovery of Potent and Orally Bioavailable N,N'-Diarylurea Antagonists for the CXCR2 Chemokine Receptor. Bioorg. Med. Chem. Lett. 2004, 14, 4375-4378. 32. McCleland, B. W.; Davis, R. S.; Palovich, M. R.; Widdowson, K. L.; Werner, M. L.; Burman, M.; Foley, J. J.; Schmidt, D. B.; Sarau, H. M.; Rogers, M.; Salyers, K. L.; Gorycki, P. D.; Roethke, T. J.; Stelman, G. J.; Azzarano, L. M.; Ward, K. W.; Busch-Petersen, J., Comparison of N,N'-Diarylsquaramides and N,N'-Diarylureas as Antagonists of the CXCR2 Chemokine Receptor. Bioorg. Med. Chem. Lett. 2007, 17, 1713-1717. 33. Merritt, J. R.; Rokosz, L. L.; Nelson, K. H., Jr.; Kaiser, B.; Wang, W.; Stauffer, T. M.; Ozgur, L. E.; Schilling, A.; Li, G.; Baldwin, J. J.; Taveras, A. G.; Dwyer, M. P.; Chao, J., Synthesis and Structure-Activity Relationships of 3,4-Diaminocyclobut-3-Ene-1,2-Dione CXCR2 Antagonists. Bioorg. Med. Chem. Lett. 2006, 16, 4107-4110. 34. Nie, H.; Widdowson, K. L.; Palovich, M. R.; Fu, W.; Elliott, J. D.; Bryan, D. L.; Burman, M.; Schmidt, D. B.; Foley, J. J.; Sarau, H. M.; Busch-Petersen, J., N,N'Diarylcyanoguanidines as Antagonists of the CXCR2 and CXCR1 Chemokine Receptors. Bioorg. Med. Chem. Lett. 2006, 16, 5513-5516. 35. Widdowson, K. L.; Elliott, J. D.; Veber, D. F.; Nie, H.; Rutledge, M. C.; McCleland, B. W.; Xiang, J. N.; Jurewicz, A. J.; Hertzberg, R. P.; Foley, J. J.; Griswold, D. E.; Martin, L.; Lee, J. M.; White, J. R.; Sarau, H. M., Evaluation of Potent and Selective Small-Molecule Antagonists for the CXCR2 Chemokine Receptor. J. Med. Chem. 2004, 47, 1319-1321. 36. Discovery Studio 2.0; Accelrys Inc, San Diego, CA, 2009. http://www.accelrys.com 37. Smellie, A.; Teig, S. L.; Towbin, P., Poling: Promoting Conformational Coverage. J Comp Chem 1995, 16, 171-187. 38. Smellie, A.; Kahn, S. D.; Teig, S. L., Analysis of Conformational Coverage. 1. Validation and Estimation of Coverage. J. Chem. Inf. Comput. Sci. 1995, 35, 285–294. 39. Smellie, A.; Kahn, S. D.; Teig, S. L., Analysis of Conformational Coverage. 2. Applications of Conformational Models. J. Chem. Inf. Comput. Sci. 1995, 35, 295–304. 40. Guner, O. F.; Henry, D. R., In Pharmacophore Perception, Development, and Use in Drug Design, Guner, O. F., Ed.; La Jolla, 2000; Chapter Metric for analyzing hit lists and pharmacophores, pp 191-212. 41. Aki, C.; Chao, J.; Ferreira, J. A.; Dwyer, M. P.; Yu, Y.; Chao, J.; Merritt, R. J.; Lai, G.; Wu, M.; Hipkin, R. W.; Fan, X.; Gonsiorek, W.; Fosseta, J.; Rindgen, D.; Fine, J.; Lundell, D.; Taveras, A. G.; Biju, P., Diaminocyclobutenediones as Potent and Orally Bioavailable CXCR2 Receptor Antagonists: Sar in the Phenolic Amide Region. Bioorg. Med. Chem. Lett. 2009, 19, 4446-4449. 42. Biju, P.; Taveras, A. G.; Dwyer, M. P.; Yu, Y.; Chao, J.; Hipkin, R. W.; Fan, X.; Rindgen, D.; Fine, J.; Lundell, D., Fluoroalkyl Alpha Side Chain Containing 3,4-DiaminoCyclobutenediones as Potent and Orally Bioavailable CXCR2-CXCR1 Dual Antagonists. Bioorg. Med. Chem. Lett. 2009, 19, 1431-1433. 43. Chao, J.; Taveras, A. G.; Chao, J.; Aki, C.; Dwyer, M.; Yu, Y.; Purakkattle, B.; Rindgen, D.; Jakway, J.; Hipkin, W.; Fosetta, J.; Fan, X.; Lundell, D.; Fine, J.; Minnicozzi, M.; Phillips, J.; Merritt, J. R., C(4)-Alkyl Substituted Furanyl Cyclobutenediones as Potent, Orally Bioavailable CXCR2 and CXCR1 Receptor Antagonists. Bioorg. Med. Chem. Lett. 2007, 17, 3778-3783.


54

Page 55 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


44. Baxter, A.; Bennion, C.; Bent, J.; Boden, K.; Brough, S.; Cooper, A.; Kinchin, E.; Kindon, N.; McInally, T.; Mortimore, M.; Roberts, B.; Unitt, J., Hit-to-Lead Studies: The Discovery of Potent, Orally Bioavailable Triazolethiol CXCR2 Receptor Antagonists. Bioorg. Med. Chem. Lett. 2003, 13, 2625-2628. 45. Hunt, F.; Austin, C.; Austin, R.; Bonnert, R.; Cage, P.; Christie, J.; Christie, M.; Dixon, C.; Hill, S.; Jewell, R.; Martin, I.; Robinson, D.; Willis, P., Sar Studies on Thiazolo[4,5D]Pyrimidine Based CXCR2 Antagonists Involving a Novel Tandem Displacement Reaction. Bioorg. Med. Chem. Lett. 2007, 17, 2731-2734. 46. Porter, D. W.; Bradley, M.; Brown, Z.; Canova, R.; Charlton, S.; Cox, B.; Hunt, P.; Kolarik, D.; Lewis, S.; O'Connor, D.; Reilly, J.; Spanka, C.; Tedaldi, L.; Watson, S. J.; Wermuth, R.; Press, N. J., The Discovery of Potent, Orally Bioavailable Pyrazolo and Triazolopyrimidine CXCR2 Receptor Antagonists. Bioorg. Med. Chem. Lett. 2014, 24, 72-76. 47. Walters, I.; Austin, C.; Austin, R.; Bonnert, R.; Cage, P.; Christie, M.; Ebden, M.; Gardiner, S.; Grahames, C.; Hill, S.; Hunt, F.; Jewell, R.; Lewis, S.; Martin, I.; David, N.; David, R., Evaluation of a Series of Bicyclic CXCR2 Antagonists. Bioorg. Med. Chem. Lett. 2008, 18, 798-803. 48. Wang, Y.; Busch-Petersen, J.; Wang, F.; Ma, L.; Fu, W.; Kerns, J. K.; Jin, J.; Palovich, M. R.; Shen, J. K.; Burman, M.; Foley, J. J.; Schmidt, D. B.; Hunsberger, G. E.; Sarau, H. M.; Widdowson, K. L., 3-Arylamino-2h-1,2,4-Benzothiadiazin-5-Ol 1,1-Dioxides as Novel and Selective CXCR2 Antagonists. Bioorg. Med. Chem. Lett. 2007, 17, 3864-3867. 49. Winters, M. P.; Crysler, C.; Subasinghe, N.; Ryan, D.; Leong, L.; Zhao, S.; Donatelli, R.; Yurkow, E.; Mazzulla, M.; Boczon, L.; Manthey, C. L.; Molloy, C.; Raymond, H.; Murray, L.; McAlonan, L.; Tomczuk, B., Carboxylic Acid Bioisosteres Acylsulfonamides, Acylsulfamides, and Sulfonylureas as Novel Antagonists of the CXCR2 Receptor. Bioorg. Med. Chem. Lett. 2008, 18, 1926-1930. 50. Yu, Y.; Dwyer, M. P.; Chao, J.; Aki, C.; Purakkattle, B.; Rindgen, D.; Bond, R.; MayerEzel, R.; Jakway, J.; Qiu, H.; Hipkin, R. W.; Fossetta, J.; Gonsiorek, W.; Bian, H.; Fan, X.; Terminelli, C.; Fine, J.; Lundell, D.; Merritt, J. R.; He, Z.; Lai, G.; Wu, M.; Taveras, A., Synthesis and Structure-Activity Relationships of Heteroaryl Substituted-3,4-Diamino-3Cyclobut-3-Ene-1,2-Dione CXCR2/CXCR1 Receptor Antagonists. Bioorg. Med. Chem. Lett. 2008, 18, 1318-1322. 51. Cutshall, N. S.; Ursino, R.; Kucera, K. A.; Latham, J.; Ihle, N. C., Nicotinamide N-Oxides as CXCR2 Antagonists. Bioorg. Med. Chem. Lett. 2001, 11, 1951-1954. 52. Liu, S.; Liu, Y.; Wang, H.; Ding, Y.; Wu, H.; Dong, J.; Wong, A.; Chen, S. H.; Li, G.; Chan, M.; Sawyer, N.; Gervais, F. G.; Henault, M.; Kargman, S.; Bedard, L. L.; Han, Y.; Friesen, R.; Lobell, R. B.; Stout, D. M., Design, Synthesis, and Evaluation of Novel 3Amino-4-Hydrazine-Cyclobut-3-Ene-1,2-Diones as Potent and Selective CXCR2 Chemokine Receptor Antagonists. Bioorg. Med. Chem. Lett. 2009, 19, 5741-5745. 53. Lu, H.; Yang, T.; Xu, Z.; Wren, P. B.; Zhang, Y.; Cai, X.; Patel, M.; Dong, K.; Zhang, Q.; Zhang, W.; Guan, X.; Xiang, J.; Elliott, J. D.; Lin, X.; Ren, F., 2-Aminopyrimidin-4(1h)One as the Novel Bioisostere of Urea: Discovery of Novel and Potent CXCR2 Antagonists. Bioorg. Med. Chem. Lett. 2014, 24, 5493-5496. 54. Maeda, D. Y.; Peck, A. M.; Schuler, A. D.; Quinn, M. T.; Kirpotina, L. N.; Wicomb, W. N.; Fan, G. H.; Zebala, J. A., Discovery of 2-[5-(4-Fluorophenylcarbamoyl)Pyridin-2Ylsulfanylmethyl]Phenylboronic Acid (Sx-517): Noncompetitive Boronic Acid Antagonist of CXCR1 and CXCR2. J. Med. Chem. 2014, 57, 8378-8397.


55


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 56 of 70

55. Porter, D. W.; Bradley, M.; Brown, Z.; Charlton, S. J.; Cox, B.; Hunt, P.; Janus, D.; Lewis, S.; Oakley, P.; O'Connor, D.; Reilly, J.; Smith, N.; Press, N. J., The Discovery of Potent, Orally Bioavailable Pyrimidine-5-Carbonitrile-6-Alkyl CXCR2 Receptor Antagonists. Bioorg. Med. Chem. Lett. 2014, 24, 3285-3290. 56. Isberg, V.; Vroling, B.; van der Kant, R.; Li, K.; Vriend, G.; Gloriam, D., Gpcrdb: An Information System for G Protein-Coupled Receptors. Nucleic Acids Res 2014, 42, D422425. 57. Fiser, A.; Sali, A., Modloop: Automated Modeling of Loops in Protein Structures. Bioinformatics 2003, 19, 2500-2501. 58. Hawkins, P. C.; Skillman, A. G.; Warren, G. L.; Ellingson, B. A.; Stahl, M. T., Conformer Generation with Omega: Algorithm and Validation Using High Quality Structures from the Protein Databank and Cambridge Structural Database. J Chem Inf Model 2010, 50, 572-584. 59. Altenburg, J. D.; Broxmeyer, H. E.; Jin, Q.; Cooper, S.; Basu, S.; Alkhatib, G., A Naturally Occurring Splice Variant of Cxcl12/Stromal Cell-Derived Factor 1 Is a Potent Human Immunodeficiency Virus Type 1 Inhibitor with Weak Chemotaxis and Cell Survival Activities. J. Virol. 2007, 81, 8140-8148. 60. Carmichael, J.; DeGraff, W. G.; Gazdar, A. F.; Minna, J. D.; Mitchell, J. B., Evaluation of a Tetrazolium-Based Semiautomated Colorimetric Assay: Assessment of Chemosensitivity Testing. Cancer Res. 1987, 47, 936-942. 61. Kilkenny, C.; Browne, W. J.; Cuthill, I. C.; Emerson, M.; Altman, D. G., Improving Bioscience Research Reporting: The Arrive Guidelines for Reporting Animal Research. PLoS biology 2010, 8, e1000412. 62. McGrath, J. C.; Drummond, G. B.; McLachlan, E. M.; Kilkenny, C.; Wainwright, C. L., Guidelines for Reporting Experiments Involving Animals: The Arrive Guidelines. Br. J. Pharmacol. 2010, 160, 1573-1576. 63. Dayam, R.; Aiello, F.; Deng, J.; Wu, Y.; Garofalo, A.; Chen, X.; Neamati, N., Discovery of Small Molecule Integrin Alphavbeta3 Antagonists as Novel Anticancer Agents. J. Med. Chem. 2006, 49, 4526-4534. 64. Güner, O. F., Pharmacophore Perception, Development and Use in Drug Design. International University Line:La Jolla, CA: 2000. 65. Vitas-M Laboratory Screening Compounds. www.vitasmlab.com 66. Enamine Screening Collection. http://www.enamine.net/ 67. Asinex Libraries. www.asinex.com 68. de Kruijf, P.; van Heteren, J.; Lim, H. D.; Conti, P. G.; van der Lee, M. M.; Bosch, L.; Ho, K. K.; Auld, D.; Ohlmeyer, M.; Smit, M. J.; Wijkmans, J. C.; Zaman, G. J.; Leurs, R., Nonpeptidergic Allosteric Antagonists Differentially Bind to the CXCR2 Chemokine Receptor. J. Pharmacol. Exp. Ther. 2009, 329, 783-790. 69. Salchow, K.; Bond, M. E.; Evans, S. C.; Press, N. J.; Charlton, S. J.; Hunt, P. A.; Bradley, M. E., A Common Intracellular Allosteric Binding Site for Antagonists of the CXCR2 Receptor. Br. J. Pharmacol. 2010, 159, 1429-1439. 70. Singh, S.; Sadanandam, A.; Varney, M. L.; Nannuru, K. C.; Singh, R. K., Small Interfering Rna-Mediated CXCR1 or CXCR2 Knock-Down Inhibits Melanoma Tumor Growth and Invasion. Int. J. Cancer 2010, 126, 328-336. 71. Goda, A. E.; Koyama, M.; Sowa, Y.; Elokely, K. M.; Yoshida, T.; Kim, B. Y.; Sakai, T., Molecular Mechanisms of the Antitumor Activity of Sb225002: A Novel Microtubule Inhibitor. Biochem. Pharmacol. 2013, 85, 1741-1752.


56

Page 57 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


72. Du, M.; Qiu, Q.; Gruslin, A.; Gordon, J.; He, M.; Chan, C. C.; Li, D.; Tsang, B. K., Sb225002 Promotes Mitotic Catastrophe in Chemo-Sensitive and -Resistant Ovarian Cancer Cells Independent of P53 Status in Vitro. PloS one 2013, 8, e54572. 73. Palm, K.; Luthman, K.; Ungell, A. L.; Strandlund, G.; Artursson, P., Correlation of Drug Absorption with Molecular Surface Properties. J. Pharm. Sci. 1996, 85, 32-39. 74. Jiang, S.; Chow, S. C.; Nicotera, P.; Orrenius, S., Intracellular Ca2+ Signals Activate Apoptosis in Thymocytes: Studies Using the Ca(2+)-Atpase Inhibitor Thapsigargin. Exp. Cell Res. 1994, 212, 84-92.


57


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 58 of 70

TOC Graphics

Discovery of Novel CXCR2 Inhibitors Using Ligand-Based Pharmacophore Models Helen Ha, Bikash Debnath, Srinivas Odde, Tim Bensman, Henry Ho, Paul M. Beringer, and Nouri Neamati


58

Page 59 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


Figure 1 A.

B.

C.

HY

8.19 Å

4.33 Å

Ar

10.86 Å 4.04 Å

5.07 Å

HBD

D.

E.

F.


6.3 Å

HBA


Figure 2

9 8

Estimated pIC50

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Page 60 of 70

7

6 Training (R=0.93)

5

Test (R=0.79)

4 4

5

6

7

Experimental pIC50


8

9

Page 61 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


Figure 3 B.

A.

C.

CX86 Fit Score: 5.81

CX25 Fit Score: 6.4

CX815 Fit Score: 6.15

E.

D.

SB 656933 Phase I (Cystic fibrosis, COPD) Fit Score: 7.4


SCH 527123 Phase II (COPD) Fit Score: 6.5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Page 62 of 70

Figure 4

External

A249

D143

Q319

T83

K320

90 D84

Y314

Internal


Page 63 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


Figure 5 A.

B. D143 Ring

A249

T83

HBA

Ring

Q319

Anion K320

Ring

Ring

D84 Y314

C.

D. D143 A249 Q319

T83 S81

K320

D84 ACS Paragon Plus Environment

Y314


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Figure 6

Scaffold A

Scaffold B

Scaffold C


Page 64 of 70

Page 65 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


Figure 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Figure 8


Page 66 of 70

Page 67 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


Figure 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Figure 10


Page 68 of 70

Page 69 of 70

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


Figure 11

A.

B. IC50

51.48

IC50

41.63 to 63.66 R square

CX4152

0.9184

*

*p = 0.0145 vs control ; N = 3 per group



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Figure 12


Page 70 of 70

Discovery of Novel CXCR2 Inhibitors Using Ligand ... - ACS Publications

Recommend Documents